OpenAI’s Financials Under Scrutiny: A Closer Look at Revenue and Compute Costs
In the realm of frenzied dealmaking and swirling rumors of an imminent IPO, OpenAI finds itself under the microscope as financial scrutiny intensifies. Recently leaked documents obtained by tech blogger Ed Zitron offer a peek into OpenAI’s financial landscape, shedding light on its revenue and compute costs over the past few years.
According to Zitron’s findings, Microsoft raked in $493.8 million in revenue share payments from OpenAI in 2024. Fast forward to the first three quarters of 2025, and that figure skyrocketed to $865.8 million, as per the documents he reviewed.
Reports suggest that OpenAI shares 20% of its revenue with Microsoft, a result of a substantial investment from the tech giant totaling over $13 billion in the AI startup. However, the arrangement gets a bit intricate as Microsoft reciprocates by sharing revenue with OpenAI, giving back around 20% of the earnings from Bing and Azure OpenAI Service. Bing, powered by OpenAI, and the OpenAI Service, offering cloud access to OpenAI’s models, play key roles in this revenue-sharing dynamic.
The leaked documents provide valuable insights into OpenAI’s financial standing, showcasing not only its revenue but also its expenditure in relation to that revenue. Going by the reported 20% revenue-share metric, OpenAI’s revenue amounted to at least $2.5 billion in 2024 and surged to $4.33 billion in the initial three quarters of 2025, with potential for even higher figures.
CEO Sam Altman’s optimistic projections hint at OpenAI’s revenue exceeding $20 billion by the end of the year, possibly reaching a staggering $100 billion by 2027. On the flip side, Zitron’s analysis indicates that OpenAI’s inference spending surged from approximately $3.8 billion in 2024 to about $8.65 billion in the first nine months of 2025. Inference, essential for running trained AI models, represents a significant chunk of OpenAI’s financial outlay.
While OpenAI primarily relies on Microsoft Azure for compute access, partnerships with other tech players like CoreWeave, Oracle, AWS, and Google Cloud have also been forged. Previous reports estimated OpenAI’s total compute spend at roughly $5.6 billion in 2024 and identified a substantial “cost of revenue” at $2.5 billion for the first half of 2025.
The intricate balance of revenue sharing and compute costs paints a complex picture of OpenAI’s financial health. With inference costs potentially outweighing revenue earnings, questions arise about the sustainability of OpenAI’s financial model and its implications for the broader AI landscape.
As the AI bubble discourse permeates industry conversations, the spotlight on OpenAI’s financial dynamics adds an intriguing layer to the narrative. With OpenAI and Microsoft remaining tight-lipped on the matter, the industry awaits further developments in this evolving financial saga.
For those with insider insights or confidential information on the AI industry’s inner workings, feel free to reach out to our reporters at rebecca.bellan@techcrunch.com or russell.brandom@techcrunch.com. Secure communication can be established via Signal at @rebeccabellan.491 and russellbrandom.49.